MDS6212 Fintech Theory and Practice: Week 5 Assignment

---

220041006 Yihang Li

Table of Contents

• 1  Questions 1:
  • 1.1  Google Trends
  • 1.2  Related queries
  • 1.3  Interest over Time¶
• 2  Questions 2:
  • • 2.0.1  Conclusion
• 3  Question 3:

Reference

Questions 1:

using Google Trends, construct a weekly index to capture political relations between U.S. and China from the US perspective, draw the variable in a graph, and discuss its time-series variation

Timeline-U.S. Relations With China

Note: the index in google trends is already weekly

!datapane login --server=https://datapane.com/ --token=<d95b6dc19091f789ac6af9a7026844a49c7b68c5>

/bin/bash: -c: line 0: syntax error near unexpected token `newline'
/bin/bash: -c: line 0: `datapane login --server=https://datapane.com/ --token=<d95b6dc19091f789ac6af9a7026844a49c7b68c5>'

import pandas as pd
from pytrends.request import TrendReq
import altair as alt
import plotly.express as px

Google Trends

use pytrends for automating downloading of reports from Google Trends.

# Take the keyword 'Trade War' as an example
pytrends = TrendReq(hl='en-US', tz=360)
keyword = 'Trade War'
pytrends.build_payload([keyword], cat=0, timeframe='today 5-y', geo='', gprop='')

Related queries

top_queries = pytrends.related_queries()[keyword]['top']

fig = px.bar(top_queries, 
      x='query',
      y='value')
fig.show()

Interest over Time¶

r_o_t = pytrends.interest_over_time().reset_index()
r_o_t.rename(columns={keyword : 'search_volume'}, inplace=True)

# We can see from the date attribute, the data is already the weekly data
r_o_t.tail(6)

	date	search_volume	isPartial
255	2020-09-20	14	False
256	2020-09-27	12	False
257	2020-10-04	15	False
258	2020-10-11	13	False
259	2020-10-18	11	False
260	2020-10-25	12	True

r_o_t_plot = alt.Chart(r_o_t).encode(x='date', y='search_volume').mark_area(line=True).interactive().properties(width=700)
r_o_t_plot

Questions 2:

using Google Trends, construct a weekly index to predict the outcome of the 2020 US presidential election, draw the variable in a graph, and tell us who are more likely to win the final election.

Reference:

US Elections in Search 2020

2020 Presidential & VP Debates

Pytrend Guidlines

kw_list = ['donald trump', 'joe biden']
kw_group = list(zip(*[iter(kw_list)]*1))
print(kw_group)

kw_grplist = [list(x) for x in kw_group]
print(kw_grplist)

#Since, every search term in a list of lists, 
#build_payload() method will have to check every query one by one through our for loop.
trendshow = TrendReq(hl='en-US', tz=360)
dict = {}
i = 0
for kw in kw_grplist:
    trendshow.build_payload(kw, timeframe = 'today 12-m', geo='US')
    dict[i] = trendshow.interest_over_time()
    i += 1

trendframe = pd.concat(dict, axis=1)
trendframe.columns = trendframe.columns.droplevel(0)
trendframe = trendframe.drop('isPartial', axis = 1)
trendframe.tail()

[('donald trump',), ('joe biden',)]
[['donald trump'], ['joe biden']]

	donald trump	joe biden
date
2020-09-27	100	100
2020-10-04	75	54
2020-10-11	41	53
2020-10-18	48	75
2020-10-25	43	65

import plotly
import plotly.graph_objects as go
from plotly.offline import download_plotlyjs, init_notebook_mode, iplot
import plotly.offline as pyo
init_notebook_mode(connected=True)

trace = [go.Scatter(
x = trendframe.index,
y = trendframe[col], name=col) for col in trendframe.columns]

data = trace
layout = go.Layout(title='Post', showlegend=True)
fig = go.Figure(data=data, layout=layout)

iplot(fig)

trace = [go.Bar(
x = trendframe.index,
y = trendframe[col], name=col) for col in trendframe.columns[0:2]]

data = trace
layout = go.Layout(title='Post', showlegend=True)
fig = go.Figure(data=data, layout=layout)

iplot(fig)

Conclusion

Joe Biden is more likely to win the final election.

Question 3:

using Baidu Index, construct an index to capture investor sentiment in the Chinese market, draw the variable in a graph, and discuss its time-series variation.

Reference:

[1] 如何利用百度指数进行数据分析？

[2] 度小满联合百度指数发布2020年一季度国民投资理财情绪指数

Here we choose the keyword '黄金', which in English means Gold, as our index to capture investor sentiment in the Chinese market

According to Baidu Search Index, the graph below is its search trend of Baidu user in China

Note:

1. Clearly, the time series variation of '黄金''s search trend is in certain period.

2. Besides, at the point of 2020-08-12, there is a peek search value that is much higher than any other points. By looking for the why this peek appeared, we got to know the reason is that the price of the Gold dropped sharply at that time.

3. Moreover, by comparing with keywords '美股' and 'a股' as follows， we can see the period is very similar. And we guess that it is due to the trade time related reason.